On the Computing and Communication Tradeoff in Reasoning-Based Multi-User Semantic Communications

TL;DR

This work addressing intermittent-link wireless networks proposes a reasoning-enabled semantic communication (SC) framework where end users leverage local computing to infer or generate data when links fail. It formulates a multi-user noncooperative game to maximize the effective semantic information, balancing semantic reliability and link disruptions, and introduces a Gaussian-process-based causal Bayesian optimization with sequential best-response updates to reach a local Nash equilibrium. The approach demonstrates substantial gains over classical communication systems, including throughput improvements (e.g., at least 16.6% in some regimes and up to 6× in low-link scenarios) and reliability boosts (up to 9× when links are sparse). The results highlight the practical value of embedding reasoning and causal concepts into SC for robust, end-to-end performance in dynamic wireless environments.

Abstract

Semantic communication (SC) is recognized as a promising approach for enabling reliable communication with minimal data transfer while maintaining seamless connectivity for a group of wireless users. Unlocking the advantages of SC for multi-user cases requires revisiting how communication and computing resources are allocated. This reassessment should consider the reasoning abilities of end-users, enabling receiving nodes to fill in missing information or anticipate future events more effectively. Yet, state-of-the-art SC systems primarily focus on resource allocation through compression based on semantic relevance, while overlooking the underlying data generation mechanisms and the tradeoff between communications and computing. Thus, they cannot help prevent a disruption in connectivity. In contrast, in this paper, a novel framework for computing and communication resource allocation is proposed that seeks to demonstrate how SC systems with reasoning capabilities at the end nodes can improve reliability in an end-to-end multi-user wireless system with intermittent communication links. Towards this end, a novel reasoning-aware SC system is proposed for enabling users to utilize their local computing resources to reason the representations when the communication links are unavailable. To optimize communication and computing resource allocation in this system, a noncooperative game is formulated among multiple users whose objective is to maximize the effective semantic information (computed as a product of reliability and semantic information) while controlling the number of semantically relevant links that are disrupted. Simulation results show that the proposed reasoning-aware SC system results in at least a $16.6\%$ enhancement in throughput and a significant improvement in reliability compared to classical communications systems that do not incorporate reasoning.

Figures (5)

• Figure 1: An illustration of a decentralized robotic terrain mapping system. Robots, operating in a shared physical environment, collaborate to build a virtual environment such as a 3D map. They exchange information through unreliable communication links, relying on local computing resources to reason about missing data. The control actions, represented by $\mathcal{A}_k$, correspond to the robots' movements and actions. The tasks in the physical space involve identifying and classifying terrain features, hazards, and potential resources. This information is then used to build a shared virtual map, accessible through an XR interface, facilitating human supervision.
• Figure 2: Loss vs iterations for $p=0.5$
• Figure 3: System throughput as a function of link availability $p$
• Figure 4: Semantic reliability as a function of link availability $p$
• Figure 5: Reasoned and communicated bits as a function of link availability $p$

Theorems & Definitions (3)

• Lemma 1
• Definition 1
• Definition 2